Validation of the Klinrisk Kidney Disease Progression Model in Individuals with IgA Nephropathy
Bibliographic record
Abstract
Background: IgA nephropathy is the most common primary glomerulonephritis worldwide, and leads to kidney failure in a significant proportion of those affected. Recent clinical trials have relied on persistence of proteinuria >1 g/day as the best predictor of poor long-term kidney outcomes, and other tools rely on detailed clinopathologic data which may not be readily available. We sought to validate the Klinrisk prediction model in a population-based cohort of all individuals aged 18 years and older with IgA nephropathy in the Canadian province of Manitoba from 2002 to 2019, inclusive. Methods: Participants were identified from the Manitoba Glomerular Diseases Registry. Klinrisk is a machine learning model constructed using random forests that uses routinely collected laboratory data to predict the risk of a composite of 40% decline in eGFR or kidney failure, defined as receipt of dialysis for at least 3 months or kidney transplantation. Discrimination was assessed using area under the receiver operating characteristic curves (AUC). Calibration was assessed using Brier scores, and calibration plots to compare predicted with observed risk. Results: A total of 230 individuals with biopsy-proven IgA nephropathy were included in the analysis. Median age at biopsy was 41.5 years, median eGFR was 51.5 mL/min/1.73m2, and median ACR was 158 mg/mmol. At 2 and 5 years, 73 and 89 individuals reached the primary outcome, respectively. Model discrimination was good, with an AUC of 0.878 (95% CI 0.830, 0.925) at 2 years and 0.827 (95% CI 0.760, 0.894) at 5 years. Calibration was appropriate, with Brier scores of 0.142 and 0.167 at 2 and 5 years. Visual inspection of the calibration plots showed under-prediction at higher levels of observed risk. Conclusion: The results demonstrate the utility of the Klinrisk model in predicting kidney failure among individuals with IgA nephropathy. Our results are limited by a small sample size and require confirmation in a separate cohort. If confirmed, Klinrisk could be implemented in a clinical setting to predict individual kidney failure risk to aid treatment decisions, or in a research setting to identify high-risk individuals for participation in clinical trials using routinely available data.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.015 | 0.014 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".